The Location Based Services have grown exponentially in the last few years with the advancement of GPS equipped devices. A lot can be explored from the GPS information, for instance it could as simple information as "where you are" to complex pattern of "busiest places in a town/city". But processing the raw GPS points is always strenuous task. As part of my course work I did a project which implemented an intuitive algorithm for stay point detection from the raw GPS log. The algorithm was from the research paper:
Yang Ye, Yu Zheng, Yukun Chen, Jianhua Feng, and Xing Xie. Mining individual life pattern based on location history. In Proceedings of the 2009 Tenth International Conference on Mobile Data Management: Systems, Services and Middleware, MDM ’09, pages 1–10, Washington, DC, USA, 2009. IEEE Computer Society.
After 500 lines of code the algorithm was up and running. I made my effort to simplify the process of converting the raw GPS log into stay points. The log that I coded for is NMEA standard and a glimpse of the log:
The way it works is:
1) upload a sample log file such as: https://www.dropbox.com/s/qjmu2w5zx67okuo/ShortLog.txt?dl=0
blog comments powered by Disqus
Yang Ye, Yu Zheng, Yukun Chen, Jianhua Feng, and Xing Xie. Mining individual life pattern based on location history. In Proceedings of the 2009 Tenth International Conference on Mobile Data Management: Systems, Services and Middleware, MDM ’09, pages 1–10, Washington, DC, USA, 2009. IEEE Computer Society.
After 500 lines of code the algorithm was up and running. I made my effort to simplify the process of converting the raw GPS log into stay points. The log that I coded for is NMEA standard and a glimpse of the log:
 |
| GPS Log Glimpse |
The GPS Log can vary from 100 traces to 10000 traces so it's important that the application is scalable, robust and yet consistent. So I have decided to write the application on a cloud platform and make available as a web service. Few jargon associated with this application are:
Stay Point: The stay point is geographical location where the user stays with in a location for substantial amount of time. Practically, these are the important locations in dially office like Gym, Home, Office, etc.
Location Sequence: The path of a location followed to reach a particular stay point. It essentially says the logical order which is followed to get to a stay point.
Stay Point: The stay point is geographical location where the user stays with in a location for substantial amount of time. Practically, these are the important locations in dially office like Gym, Home, Office, etc.
Location Sequence: The path of a location followed to reach a particular stay point. It essentially says the logical order which is followed to get to a stay point.
Finally, I present you the application:
WebApp: pseudoajgeostay.herokuapp.com
1) upload a sample log file such as: https://www.dropbox.com/s/qjmu2w5zx67okuo/ShortLog.txt?dl=0
2) It will process the log and generate stay points, sequence, unique labels and corresponding coordinates. The result will give the stay points in real map environment with varying transparency corresponding to stay point population.
P.S: Try running from source as the web service is still in process of updating. I will try to add few features, if you have any suggestions drop in comments.